Machine learning assisted segmentation of scanning electron microscopy images of organic-rich shales with feature extraction and feature ranking

Abstract

Abstract Scanning electron microscopy (SEM) image captures the high-resolution microstructure of a material. Segmentation of SEM image delineates and separates distinct components in the image. We tested an automated SEM-image segmentation workflow involving feature extraction followed by machine learning. For each pixel in the SEM image, 16 features are generated and then fed to a random forest classifier to determine the presence of four rock components, namely, (1) pore/crack; (2) rock matrix including clay, calcite, and quartz; (3) pyrite; and (4) organic/kerogen components. The most important features for the desired segmentation are wavelet transforms, Gaussian blur, and difference of Gaussians. The random forest classifier was trained using only 705, 2074, 17,373, and 15,000 pixels for the pore/crack, organic/kerogen, rock matrix, and pyrite components, respectively. For purposes of validation of the segmentation method, 5121, 4725, 4815, and 4775 pixels were used for the pore/crack, organic/kerogen, rock matrix, and pyrite components, respectively. The performance of the trained classifier, quantified in terms of overall F1 score, on the validation dataset was higher than 0.9. The newly developed method is significantly more reliable and robust in comparison with the popular histogram thresholding and object-based segmentation methods, especially for the matrix and pyrite components in shale samples.